Bioinformatics
We have worked on a machine learning system, based in Tokyo (www.nii.ac.jp - Japan)
and allowing to extract new kind of rules from a given database. This system CF-induction,
more powerful from a theoretical viewpoint than classical Inductive system,
We focus on the web interface and our first application concern is biological data.
We work with the Laboratory of Automatism and Analysis of Systems (www.laas.fr) LAAS. Our aim with this project is to apply techniques issued from the field of Inductive Logic Programming to a set of data coming from the Biotech. lab. of INSA Toulouse (France). . These data are obtained from the continuous observation of a biological process, namely yeast fermentation. Such a dynamic process is of very high importance because it can lead to high added value end-products like vitamins, antibiotics, etc. Furthermore, today, some governments are encouraging research about the production of ethanol as a alternative to standard oil-based vehicules's carburant. It makes no doubt that, in a sustainable development perspective, this research topics has to be deeply investigated. So, we understand that the more the percentage of end ethanol, the more the fermentation process is successful. Unfortunately, today, micro-biologists are not able to modelise the process in order to dynamically tune the relevant parameters insuring a high percentage of ethanol output.
We have published diverse papers on that work in ICEIS2005 (
ICEIS 2005 - 7th International Conference on Enterprise Information Systems,
Intern. Conf On ILP - 2006,
)
> >>More
|
